At HumanX Amsterdam, AI Startups Target Production Deployment Gap
SiliconANGLE reports from HumanX Amsterdam 2026, where vendors showcased tools for AI lineage, legacy maintenance, tribal knowledge, small language models, agent testing and agentic governance.
SiliconANGLE said it attended HumanX Amsterdam 2026 after AGNTCon + MCPCon Europe, aiming to uncover innovation stories across the software landscape. The report said that despite heavy hype around AI, it found several vendors laser-focused on delivering differentiated business value and helping customers advance AI initiatives. The publication described the vendors as working to close the gap between 'eyes'—the desire for AI—and 'stomachs'—the ability to deploy real solutions into production. SiliconANGLE said it selected 10 vendors for its list; among those detailed were the following.
8wave Oy offers a platform that connects an organization’s AI activities by tracking the lineage of underlying data feeding various AI applications. SiliconANGLE described the result as 'enterprise resource planning or ERP for AI operations,' connecting AI activities to key performance indicators important to the business. The company updated the notion of data lineage to 'AI lineage,' linking underlying data to core business drivers via AI.
Bosun B.V. uses AI to automate software maintenance across legacy codebases, supporting security fixes, dependency updates and some legacy migration activities. The company tracks the behavior of each legacy asset with a parity file that lets organizations compare the behavior of original and updated applications. SiliconANGLE said the parity file tracks each legacy application’s behavior through inputs and outputs in a human-readable and modifiable format, enabling patches and updates while lowering the risk of breaking something.
Clarifeye SAS offers an 'AI consultant' that interviews experts across an organization to collect tribal knowledge locked in people’s heads. It then combines that knowledge with other sources of information to help companies build process maps, business rules and playbooks that accurately represent the actual behaviors of the business. SiliconANGLE said Clarifeye complements digital sources such as documents, databases and applications with hard-to-reach tribal knowledge about how businesses actually operate.
Distil labs GmbH simplifies the creation of small, or specialized, language models (SLMs), enabling customers to dramatically lower token costs compared with large language models (LLMs). Distil iteratively fine-tunes an existing model, leveraging corporate and synthetic data to train and deploy customized SLMs that meet specific business requirements. SiliconANGLE said the company’s automated, iterative approach works for any use case and any customer, unlike prebuilt SLMs on the market that can only deliver specific, fixed value propositions.
Reasoning Engines B.V., doing business as LangWatch, provides simulation-based AI agent testing and evaluation. At the core of the LangWatch offering are agents that simulate human behavior as they interact with an application under test. LangWatch can then iteratively write and improve tests based on its output and can also use these tests as agent evaluations, or evals. SiliconANGLE said LangWatch is an early adopter of the new Jev AI model from TypeSafe AI Inc., which promises to dramatically accelerate agentic evals and associated LLM-as-a-Judge capabilities, resolving what the report described as the primary roadblock to agentic evaluation.
Orq.ai Holding B.V. offers an agentic AI development and deployment platform that combines centralized agentic governance with decentralized, domain-specific agentic delivery capabilities. The centralized portion serves as an agentic center of excellence, or CoE, offering organization-wide governance including policies, guardrails and audit trails. Under the CoE’s purview, individual teams can leverage any agentic technologies to build agents or applications they require while remaining compliant. SiliconANGLE said Orq.ai leverages platform engineering best practices in its dual-platform architecture, where the platform engineering team establishes the golden path of favored tooling for distributed teams, balancing centralized governance priorities with diverse business needs.